suspend/resume fails on second attempt in LNXVIDEO:00

suspend/resume fails on second attempt in LNXVIDEO:00

[Brian J. Murrell]

I'm trying to debug a resume problem on a compaq presario C700 laptop on 
the Ubuntu 2.6.24-22-generic kernel.

I've used the /sys/power/pm_trace facility to debug the problem to the
following:

[   31.212062]   Magic number: 0:716:374
[   31.212115]   hash matches /build/buildd/linux-2.6.24/drivers/base/power/main.c:97
[   31.212255]   hash matches device LNXVIDEO:00

Not sure where to go from here though.  Devices in this machine are:

00:00.0 Host bridge: Intel Corporation Mobile PM965/GM965/GL960 Memory Controller Hub (rev 03)
00:02.0 VGA compatible controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (rev 03)
00:02.1 Display controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (rev 03)
00:1b.0 Audio device: Intel Corporation 82801H (ICH8 Family) HD Audio Controller (rev 03)
00:1c.0 PCI bridge: Intel Corporation 82801H (ICH8 Family) PCI Express Port 1 (rev 03)
00:1d.0 USB Controller: Intel Corporation 82801H (ICH8 Family) USB UHCI Controller #1 (rev 03)
00:1d.1 USB Controller: Intel Corporation 82801H (ICH8 Family) USB UHCI Controller #2 (rev 03)
00:1d.2 USB Controller: Intel Corporation 82801H (ICH8 Family) USB UHCI Controller #3 (rev 03)
00:1d.7 USB Controller: Intel Corporation 82801H (ICH8 Family) USB2 EHCI Controller #1 (rev 03)
00:1e.0 PCI bridge: Intel Corporation 82801 Mobile PCI Bridge (rev f3)
00:1f.0 ISA bridge: Intel Corporation 82801HEM (ICH8M) LPC Interface Controller (rev 03)
00:1f.1 IDE interface: Intel Corporation 82801HBM/HEM (ICH8M/ICH8M-E) IDE Controller (rev 03)
00:1f.2 SATA controller: Intel Corporation 82801HBM/HEM (ICH8M/ICH8M-E) SATA AHCI Controller (rev 03)
00:1f.3 SMBus: Intel Corporation 82801H (ICH8 Family) SMBus Controller (rev 03)
01:00.0 Ethernet controller: Atheros Communications Inc. AR242x 802.11abg Wireless PCI Express Adapter (rev 01)
02:01.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139/8139C/8139C+ (rev 10)

These symptoms mirror a lot of other reports with this machine in that 
the first suspend/resume cycle works but a second attempts (apparently) 
suspends fine but fails on resume.  The backlight on the video does not 
come back and neither does the wireless connection, so I'm not positive 
the problems are purely video related.

How can I proceed from here to further debug what kind of workarounds or
solutions I might be able to use to get reliable suspend/resume 
functionality?

Thanx,
b.

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Justin P. Mattock]

Brian J. Murrell wrote:
> I'm trying to debug a resume problem on a compaq presario C700 laptop on 
> the Ubuntu 2.6.24-22-generic kernel.
>
> I've used the /sys/power/pm_trace facility to debug the problem to the
> following:
>
> [   31.212062]   Magic number: 0:716:374
> [   31.212115]   hash matches /build/buildd/linux-2.6.24/drivers/base/power/main.c:97
> [   31.212255]   hash matches device LNXVIDEO:00
>
> Not sure where to go from here though.  Devices in this machine are:
>
> 00:00.0 Host bridge: Intel Corporation Mobile PM965/GM965/GL960 Memory Controller Hub (rev 03)
> 00:02.0 VGA compatible controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (rev 03)
> 00:02.1 Display controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (rev 03)
> 00:1b.0 Audio device: Intel Corporation 82801H (ICH8 Family) HD Audio Controller (rev 03)
> 00:1c.0 PCI bridge: Intel Corporation 82801H (ICH8 Family) PCI Express Port 1 (rev 03)
> 00:1d.0 USB Controller: Intel Corporation 82801H (ICH8 Family) USB UHCI Controller #1 (rev 03)
> 00:1d.1 USB Controller: Intel Corporation 82801H (ICH8 Family) USB UHCI Controller #2 (rev 03)
> 00:1d.2 USB Controller: Intel Corporation 82801H (ICH8 Family) USB UHCI Controller #3 (rev 03)
> 00:1d.7 USB Controller: Intel Corporation 82801H (ICH8 Family) USB2 EHCI Controller #1 (rev 03)
> 00:1e.0 PCI bridge: Intel Corporation 82801 Mobile PCI Bridge (rev f3)
> 00:1f.0 ISA bridge: Intel Corporation 82801HEM (ICH8M) LPC Interface Controller (rev 03)
> 00:1f.1 IDE interface: Intel Corporation 82801HBM/HEM (ICH8M/ICH8M-E) IDE Controller (rev 03)
> 00:1f.2 SATA controller: Intel Corporation 82801HBM/HEM (ICH8M/ICH8M-E) SATA AHCI Controller (rev 03)
> 00:1f.3 SMBus: Intel Corporation 82801H (ICH8 Family) SMBus Controller (rev 03)
> 01:00.0 Ethernet controller: Atheros Communications Inc. AR242x 802.11abg Wireless PCI Express Adapter (rev 01)
> 02:01.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139/8139C/8139C+ (rev 10)
>
> These symptoms mirror a lot of other reports with this machine in that 
> the first suspend/resume cycle works but a second attempts (apparently) 
> suspends fine but fails on resume.  The backlight on the video does not 
> come back and neither does the wireless connection, so I'm not positive 
> the problems are purely video related.
>
> How can I proceed from here to further debug what kind of workarounds or
> solutions I might be able to use to get reliable suspend/resume 
> functionality?
>
> Thanx,
> b.
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>
>   
not too sure what the situation is for you, but
if  you're deciding to stay with 2.6.24.* then
look through the commits to see if this has been resolved.
if not then load the latest git to see if this has been committed
for the fix. if it's just debugging then hmm.. maybe ssh
(but you don't have intranet upon wakeup); maybe usb
or some other hard wired mechanism?

regards;

Justin P. Mattock



 

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Brian J. Murrell]

On Sat, 13 Dec 2008 22:43:09 -0800, Justin P. Mattock wrote:

> not too sure what the situation is for you, but if  you're deciding to
> stay with 2.6.24.* then look through the commits to see if this has been
> resolved. if not then load the latest git to see if this has been
> committed for the fix.

Well, I've just built and installed 2.6.28-rc8 and the problem persists.

> if it's just debugging then hmm.. maybe ssh (but
> you don't have intranet upon wakeup);

Right.  It's for all intents and purposes "dead" on resume.  Even 
CAPSLOCK on the console does not active the LED.

> maybe usb or some other hard wired
> mechanism?

I'd hoped to do serial console, but sadly, this laptop, like many today 
doesn't have even a single serial port.  I had considered a USB serial 
and making it console (if even possible) but now I am just spending more 
money trying to get this stupid laptop working.  ~sigh~

b.

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Justin P. Mattock]

Brian J. Murrell wrote:
> On Sat, 13 Dec 2008 22:43:09 -0800, Justin P. Mattock wrote:
>
>   
>> not too sure what the situation is for you, but if  you're deciding to
>> stay with 2.6.24.* then look through the commits to see if this has been
>> resolved. if not then load the latest git to see if this has been
>> committed for the fix.
>>     
>
> Well, I've just built and installed 2.6.28-rc8 and the problem persists.
>
>   
>> if it's just debugging then hmm.. maybe ssh (but
>> you don't have intranet upon wakeup);
>>     
>
> Right.  It's for all intents and purposes "dead" on resume.  Even 
> CAPSLOCK on the console does not active the LED.
>
>   
>> maybe usb or some other hard wired
>> mechanism?
>>     
>
> I'd hoped to do serial console, but sadly, this laptop, like many today 
> doesn't have even a single serial port.  I had considered a USB serial 
> and making it console (if even possible) but now I am just spending more 
> money trying to get this stupid laptop working.  ~sigh~
>
> b.
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>
>   
 From over here with a macbook(no serial port as well)
making debugging intricate to the extent.
With running radeon it's weired,
 if I do an echo mem > /sys/power/state
upon wakeup I'll have a black screen,(intranet works)
but if I do a  s2ram "poof" like magic it works.
if using (sorry for the word) fglrx, echo mem
works. haven't tried s2ram with that though.
Aside note glad to see you're running the latest kernel.
for the past week or so there was posts about suspend
but haven't kept up(was busy looking at SELinux refpolicy UBAC)
maybe there was some talks or even some patches submitted
for what you're dealing with?

regards;

Justin P. Mattock

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Pavel Machek]

On Sun 2008-12-14 05:06:01, Brian J. Murrell wrote:
> I'm trying to debug a resume problem on a compaq presario C700 laptop on 
> the Ubuntu 2.6.24-22-generic kernel.
> 
> I've used the /sys/power/pm_trace facility to debug the problem to the
> following:
> 
> [   31.212062]   Magic number: 0:716:374
> [   31.212115]   hash matches /build/buildd/linux-2.6.24/drivers/base/power/main.c:97
> [   31.212255]   hash matches device LNXVIDEO:00
> 
> Not sure where to go from here though.  Devices in this machine are:

Sounds like acpi video driver to me. So you may want to rmmod it
before suspend at see what happens. (and cc linux-acpi).

-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Brian J. Murrell]

On Mon, 15 Dec 2008 16:56:55 +0100, Pavel Machek wrote:
 
> Sounds like acpi video driver to me. So you may want to rmmod it before
> suspend at see what happens. (and cc linux-acpi).

Please forgive my ignorance.  Which driver would that be?  I currently
have:

$ lsmod
Module                  Size  Used by
af_packet              25088  2 
i915                   64004  2 
drm                    93736  3 i915
rfcomm                 41744  2 
l2cap                  29312  13 rfcomm
bluetooth              62436  4 rfcomm,l2cap
ppdev                  15620  0 
ipv6                  254836  24 
acpi_cpufreq           16268  1 
cpufreq_conservative    14472  0 
cpufreq_powersave       9856  0 
cpufreq_ondemand       15244  1 
cpufreq_userspace      11268  0 
cpufreq_stats          13188  0 
freq_table             12672  3 acpi_cpufreq,cpufreq_ondemand,cpufreq_stats
container              11648  0 
pci_slot               12680  0 
sbs                    18952  0 
sbshc                  13568  1 sbs
iptable_filter         11008  0 
ip_tables              19216  1 iptable_filter
x_tables               22916  1 ip_tables
parport_pc             40356  0 
lp                     17156  0 
parport                42348  3 ppdev,parport_pc,lp
joydev                 17856  0 
snd_hda_intel         416052  0 
snd_seq_dummy          11012  0 
snd_pcsp               17832  0 
snd_seq_oss            37248  0 
snd_pcm_oss            44928  0 
snd_mixer_oss          22272  1 snd_pcm_oss
psmouse                49168  0 
snd_seq_midi           14464  0 
snd_rawmidi            29184  1 snd_seq_midi
snd_seq_midi_event     15232  2 snd_seq_oss,snd_seq_midi
evdev                  17696  6 
snd_seq                55984  6 snd_seq_dummy,snd_seq_oss,snd_seq_midi,snd_seq_midi_event
snd_pcm                81156  3 snd_hda_intel,snd_pcsp,snd_pcm_oss
serio_raw              13444  0 
snd_timer              28808  2 snd_seq,snd_pcm
snd_seq_device         15244  5 snd_seq_dummy,snd_seq_oss,snd_seq_midi,snd_rawmidi,snd_seq
iTCO_wdt               19236  0 
iTCO_vendor_support    11908  1 iTCO_wdt
video                  25360  0 
output                 11136  1 video
snd                    61476  10 snd_hda_intel,snd_pcsp,snd_seq_oss,snd_pcm_oss,snd_mixer_oss,snd_rawmidi,snd_seq,snd_pcm,snd_timer,snd_seq_device
soundcore              15328  1 snd
ac                     12420  0 
snd_page_alloc         17160  2 snd_hda_intel,snd_pcm
wmi                    14888  0 
battery                18436  0 
shpchp                 39828  0 
pci_hotplug            34976  1 shpchp
intel_agp              33596  1 
agpgart                41928  3 drm,intel_agp
button                 14480  0 
ext3                  130568  1 
jbd                    53908  1 ext3
mbcache                16004  1 ext3
loop                   22540  2 
usb_storage            78656  0 
libusual               30356  1 usb_storage
sd_mod                 41240  2 
crc_t10dif             10112  1 sd_mod
sr_mod                 21956  0 
cdrom                  42272  1 sr_mod
sg                     36148  0 
ahci                   37260  1 
ata_piix               29700  0 
pata_acpi              12288  0 
ata_generic            13060  0 
8139too                31744  0 
8139cp                 27776  0 
mii                    13440  2 8139too,8139cp
libata                176160  4 ahci,ata_piix,pata_acpi,ata_generic
scsi_mod              156948  5 usb_storage,sd_mod,sr_mod,sg,libata
ehci_hcd               41996  0 
uhci_hcd               30352  0 
usbcore               149392  5 usb_storage,libusual,ehci_hcd,uhci_hcd
thermal                23708  0 
processor              49708  4 acpi_cpufreq,thermal
fan                    12676  0 
fuse                   58780  3 

Do you mean the i915 driver?  I don't think I will be able to rmmod that
one given the number of references on it.

Thanx!

b.

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Alexey Starikovskiy]

Brian J. Murrell wrote:
> On Mon, 15 Dec 2008 16:56:55 +0100, Pavel Machek wrote:
>  
>   
>> Sounds like acpi video driver to me. So you may want to rmmod it before
>> suspend at see what happens. (and cc linux-acpi).
>>     
>
> Please forgive my ignorance.  Which driver would that be?  I currently
> have:
>
> $ lsmod
> Module                  Size  Used by
> af_packet              25088  2 
> i915                   64004  2 
> drm                    93736  3 i915
> rfcomm                 41744  2 
> l2cap                  29312  13 rfcomm
> bluetooth              62436  4 rfcomm,l2cap
> ppdev                  15620  0 
> ipv6                  254836  24 
> acpi_cpufreq           16268  1 
> cpufreq_conservative    14472  0 
> cpufreq_powersave       9856  0 
> cpufreq_ondemand       15244  1 
> cpufreq_userspace      11268  0 
> cpufreq_stats          13188  0 
> freq_table             12672  3 acpi_cpufreq,cpufreq_ondemand,cpufreq_stats
> container              11648  0 
> pci_slot               12680  0 
> sbs                    18952  0 
> sbshc                  13568  1 sbs
> iptable_filter         11008  0 
> ip_tables              19216  1 iptable_filter
> x_tables               22916  1 ip_tables
> parport_pc             40356  0 
> lp                     17156  0 
> parport                42348  3 ppdev,parport_pc,lp
> joydev                 17856  0 
> snd_hda_intel         416052  0 
> snd_seq_dummy          11012  0 
> snd_pcsp               17832  0 
> snd_seq_oss            37248  0 
> snd_pcm_oss            44928  0 
> snd_mixer_oss          22272  1 snd_pcm_oss
> psmouse                49168  0 
> snd_seq_midi           14464  0 
> snd_rawmidi            29184  1 snd_seq_midi
> snd_seq_midi_event     15232  2 snd_seq_oss,snd_seq_midi
> evdev                  17696  6 
> snd_seq                55984  6 snd_seq_dummy,snd_seq_oss,snd_seq_midi,snd_seq_midi_event
> snd_pcm                81156  3 snd_hda_intel,snd_pcsp,snd_pcm_oss
> serio_raw              13444  0 
> snd_timer              28808  2 snd_seq,snd_pcm
> snd_seq_device         15244  5 snd_seq_dummy,snd_seq_oss,snd_seq_midi,snd_rawmidi,snd_seq
> iTCO_wdt               19236  0 
> iTCO_vendor_support    11908  1 iTCO_wdt
> video                  25360  0 
>   
this one ^^^^^
> output                 11136  1 video
> snd                    61476  10 snd_hda_intel,snd_pcsp,snd_seq_oss,snd_pcm_oss,snd_mixer_oss,snd_rawmidi,snd_seq,snd_pcm,snd_timer,snd_seq_device
> soundcore              15328  1 snd
> ac                     12420  0 
> snd_page_alloc         17160  2 snd_hda_intel,snd_pcm
> wmi                    14888  0 
> battery                18436  0 
> shpchp                 39828  0 
> pci_hotplug            34976  1 shpchp
> intel_agp              33596  1 
> agpgart                41928  3 drm,intel_agp
> button                 14480  0 
> ext3                  130568  1 
> jbd                    53908  1 ext3
> mbcache                16004  1 ext3
> loop                   22540  2 
> usb_storage            78656  0 
> libusual               30356  1 usb_storage
> sd_mod                 41240  2 
> crc_t10dif             10112  1 sd_mod
> sr_mod                 21956  0 
> cdrom                  42272  1 sr_mod
> sg                     36148  0 
> ahci                   37260  1 
> ata_piix               29700  0 
> pata_acpi              12288  0 
> ata_generic            13060  0 
> 8139too                31744  0 
> 8139cp                 27776  0 
> mii                    13440  2 8139too,8139cp
> libata                176160  4 ahci,ata_piix,pata_acpi,ata_generic
> scsi_mod              156948  5 usb_storage,sd_mod,sr_mod,sg,libata
> ehci_hcd               41996  0 
> uhci_hcd               30352  0 
> usbcore               149392  5 usb_storage,libusual,ehci_hcd,uhci_hcd
> thermal                23708  0 
> processor              49708  4 acpi_cpufreq,thermal
> fan                    12676  0 
> fuse                   58780  3 
>
> Do you mean the i915 driver?  I don't think I will be able to rmmod that
> one given the number of references on it.
>
> Thanx!
>
> b.
>
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>   

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Rafael J. Wysocki]

On Monday, 15 of December 2008, Brian J. Murrell wrote:
> On Mon, 15 Dec 2008 20:33:44 +0300, Alexey Starikovskiy wrote:
> 
> >> video                  25360  0
> >>   
> > this one ^^^^^
> 
> Ahhh.  Got it.
> 
> Unfortunately that was of no help.  Still only a single suspend/resume 
> cycle worked.  I removed it before the first cycle, (and made sure it was 
> not loaded before the second) for the record.

Does it also happen in the minimal configuration (ie. with init=/bin/bash)?

Rafael

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Rafael J. Wysocki]

On Monday, 15 of December 2008, Brian J. Murrell wrote:
> On Mon, 15 Dec 2008 21:19:34 +0100, Rafael J. Wysocki wrote:
> > Does it also happen in the minimal configuration (ie. with
> > init=/bin/bash)?
> 
> Now, that's an interesting result.  Using the above, pm-suspend does 
> suspend fine and hitting the flashing power button will resume fine with 
> the exception of the screen is still black.  That is backlight problems 
> isn't it?  I can see by disk light activity though that it's actually 
> still alive though.

No, most probably your graphics adapter is not handled correctly in this case,
because it needs some special user space quirks, which are done by hald that is
not running.

You can try to use the s2ram binary (http://en.opensuse.org/s2ram) to work
around this issue.

Anyway, if you suspend for the second time from this state, does it resume?

Rafael

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Rafael J. Wysocki]

On Tuesday, 16 of December 2008, Brian J. Murrell wrote:
> On Tue, 16 Dec 2008 00:02:43 +0100, Rafael J. Wysocki wrote:
> > 
> > No, most probably your graphics adapter is not handled correctly in this
> > case, because it needs some special user space quirks, which are done by
> > hald that is not running.
> 
> Hrm.  OK.
>  
> > You can try to use the s2ram binary (http://en.opensuse.org/s2ram) to
> > work around this issue.
> 
> That sounds like I am setting for something less than optimal.  Is that 
> the case?  Why would I want to use what I've been trying to get working 
> rather than s2ram?  Maybe put another way, why isn't everyone using s2ram?

Because the hal-driven whitelist allows us to match systems in a more detailed
way.  If your system is in the s2ram whitelist already and works with it, there
shouldn't be any difference.

Still, I was only considering that as a debugging aid in your case, because
s2ram works in the minimal configuration, while hald doesn't.

> Ultimately I'm going to want to be able to suspend as well as hibernate, 
> for whatever that's worth.
> 
> > Anyway, if you suspend for the second time from this state, does it
> > resume?
> 
> No.  It fails just as miserably.  :-(

That's a bit of new information.

What's the list of modules loaded in the minimal configuration?

Rafael

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Brian J. Murrell]

On Tue, 16 Dec 2008 16:25:37 +0100, Rafael J. Wysocki wrote:

> Because the hal-driven whitelist allows us to match systems in a more
> detailed way.  If your system is in the s2ram whitelist already and
> works with it, there shouldn't be any difference.

Unfortunately, Ubuntu in their "wisdom" removed s2ram and claim "s2both" 
supersedes it in that it does both suspend to disk and suspend to ram and 
then suspends, but if your battery dies while suspended you can still 
resume from the suspend to disk.
 
> Still, I was only considering that as a debugging aid in your case,
> because s2ram works in the minimal configuration, while hald doesn't.

Right.  I will give s2both a try and see what happens... no joy.  It 
fails to even run without the swap device mounted, and my swap device is 
not big enough to hibernate to anyway.  Damn you Ubuntu.

> That's a bit of new information.
> 
> What's the list of modules loaded in the minimal configuration?

Module                  Size  Used by
ext3                  130568  1 
jbd                    53908  1 ext3
mbcache                16004  1 ext3
loop                   22540  2 
usb_storage            78656  0 
libusual               30356  1 usb_storage
sd_mod                 41240  2 
crc_t10dif             10112  1 sd_mod
sr_mod                 21956  0 
cdrom                  42272  1 sr_mod
sg                     36148  0 
ahci                   37260  1 
ata_piix               29700  0 
pata_acpi              12288  0 
ata_generic            13060  0 
libata                176160  4 ahci,ata_piix,pata_acpi,ata_generic
scsi_mod              156948  5 usb_storage,sd_mod,sr_mod,sg,libata
8139too                31744  0 
8139cp                 27776  0 
mii                    13440  2 8139too,8139cp
ehci_hcd               41996  0 
uhci_hcd               30352  0 
usbcore               149392  5 usb_storage,libusual,ehci_hcd,uhci_hcd
thermal                23708  0 
processor              49708  3 thermal
fan                    12676  0 
fuse                   58780  3 

Cheers,
b.

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Rafael J. Wysocki]

On Tuesday, 16 of December 2008, Brian J. Murrell wrote:
> On Tue, 16 Dec 2008 16:25:37 +0100, Rafael J. Wysocki wrote:
> 
> > Because the hal-driven whitelist allows us to match systems in a more
> > detailed way.  If your system is in the s2ram whitelist already and
> > works with it, there shouldn't be any difference.
> 
> Unfortunately, Ubuntu in their "wisdom" removed s2ram and claim "s2both" 
> supersedes it in that it does both suspend to disk and suspend to ram and 
> then suspends, but if your battery dies while suspended you can still 
> resume from the suspend to disk.

Hm, this really isn't reasonable.  s2both saves the image, which takes quite
a lot of time, while s2ram doesn't.  Also, s2both is not really ACPI-compliant.

> > Still, I was only considering that as a debugging aid in your case,
> > because s2ram works in the minimal configuration, while hald doesn't.
> 
> Right.  I will give s2both a try and see what happens... no joy.  It 
> fails to even run without the swap device mounted, and my swap device is 
> not big enough to hibernate to anyway.  Damn you Ubuntu.

You can compile s2ram from sources, it's not too difficult.

> > That's a bit of new information.
> > 
> > What's the list of modules loaded in the minimal configuration?
> 
> Module                  Size  Used by
> ext3                  130568  1 
> jbd                    53908  1 ext3
> mbcache                16004  1 ext3
> loop                   22540  2 
> usb_storage            78656  0 
> libusual               30356  1 usb_storage
> sd_mod                 41240  2 
> crc_t10dif             10112  1 sd_mod
> sr_mod                 21956  0 
> cdrom                  42272  1 sr_mod
> sg                     36148  0 
> ahci                   37260  1 
> ata_piix               29700  0 
> pata_acpi              12288  0 
> ata_generic            13060  0 
> libata                176160  4 ahci,ata_piix,pata_acpi,ata_generic
> scsi_mod              156948  5 usb_storage,sd_mod,sr_mod,sg,libata
> 8139too                31744  0 
> 8139cp                 27776  0 
> mii                    13440  2 8139too,8139cp
> ehci_hcd               41996  0 
> uhci_hcd               30352  0 
> usbcore               149392  5 usb_storage,libusual,ehci_hcd,uhci_hcd
> thermal                23708  0 
> processor              49708  3 thermal
> fan                    12676  0 
> fuse                   58780  3 

Hm.  Is that all loaded when you boot with init=/bin/bash (by which I mean the
minimal config)?

Rafael

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Brian J. Murrell]

On Tue, 16 Dec 2008 17:13:08 +0100, Rafael J. Wysocki wrote:
> 
> Hm, this really isn't reasonable.

Many people agree with you.  :-)  There is a long bug open with ubuntu to 
restore this s2ram.  They seem to be ignoring it though.

> s2both saves the image, which takes
> quite a lot of time, while s2ram doesn't.

Indeed.
 
> You can compile s2ram from sources, it's not too difficult.

Yeah.  I will probably just grab the debian package and dpkg-buildpackage 
it on my Ubuntu system.
 
> Hm.  Is that all loaded when you boot with init=/bin/bash (by which I
> mean the minimal config)?

It sure is.  Most assuredly by the initrd.  Debian/Ubuntu systems seem to 
put everything but the kitchen sink into their initrds.  Which on the one 
hand makes them pretty portable and resilient, but on the other, yes, 
bloated.

b.

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Rafael J. Wysocki]

On Tuesday, 16 of December 2008, Brian J. Murrell wrote:
> On Tue, 16 Dec 2008 17:13:08 +0100, Rafael J. Wysocki wrote:
> > 
> > Hm, this really isn't reasonable.
> 
> Many people agree with you.  :-)  There is a long bug open with ubuntu to 
> restore this s2ram.  They seem to be ignoring it though.
> 
> > s2both saves the image, which takes
> > quite a lot of time, while s2ram doesn't.
> 
> Indeed.
>  
> > You can compile s2ram from sources, it's not too difficult.
> 
> Yeah.  I will probably just grab the debian package and dpkg-buildpackage 
> it on my Ubuntu system.
>  
> > Hm.  Is that all loaded when you boot with init=/bin/bash (by which I
> > mean the minimal config)?
> 
> It sure is.  Most assuredly by the initrd.  Debian/Ubuntu systems seem to 
> put everything but the kitchen sink into their initrds.  Which on the one 
> hand makes them pretty portable and resilient, but on the other, yes, 
> bloated.

Well, please remove as many modules as you can (using rmmod) and see if your
second resume still fails.

Thanks,
Rafael

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Matthew Garrett]

On Tue, Dec 16, 2008 at 05:13:08PM +0100, Rafael J. Wysocki wrote:
> On Tuesday, 16 of December 2008, Brian J. Murrell wrote:
> > On Tue, 16 Dec 2008 16:25:37 +0100, Rafael J. Wysocki wrote:
> > 
> > > Because the hal-driven whitelist allows us to match systems in a more
> > > detailed way.  If your system is in the s2ram whitelist already and
> > > works with it, there shouldn't be any difference.
> > 
> > Unfortunately, Ubuntu in their "wisdom" removed s2ram and claim "s2both" 
> > supersedes it in that it does both suspend to disk and suspend to ram and 
> > then suspends, but if your battery dies while suspended you can still 
> > resume from the suspend to disk.

I seem to be missing the original message, but: no, s2ram doesn't exist 
in Ubuntu because pm-utils handles that role.

-- 
Matthew Garrett | mjg59@srcf.ucam.org

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Matthew Garrett]

On Tue, Dec 16, 2008 at 06:07:15PM +0000, Brian J. Murrell wrote:
> On Tue, 16 Dec 2008 16:37:46 +0000, Matthew Garrett wrote:
> > 
> > I seem to be missing the original message, but: no, s2ram doesn't exist
> > in Ubuntu because pm-utils handles that role.
> 
> Well, that's *their* argument but lots of people think otherwise.  There 
> appears to be no shortage of people for whom pm-utils does NOT work yet 
> s2ram does.  https://bugs.launchpad.net/ubuntu/+source/uswsusp/+bug/134238

The right answer to "This piece of software contains bugs" is not 
"Provide two pieces of software with the same features but different 
bugs". The only real functional difference between the two is that s2ram 
has its own internal whitelist rather than just using the hal entries. 
Suse have normally been good at making sure that they're synced across, 
but if that's not happening you probably want to poke them.

-- 
Matthew Garrett | mjg59@srcf.ucam.org

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Pavel Machek]

On Tue 2008-12-16 20:53:35, Matthew Garrett wrote:
> On Tue, Dec 16, 2008 at 06:07:15PM +0000, Brian J. Murrell wrote:
> > On Tue, 16 Dec 2008 16:37:46 +0000, Matthew Garrett wrote:
> > > 
> > > I seem to be missing the original message, but: no, s2ram doesn't exist
> > > in Ubuntu because pm-utils handles that role.
> > 
> > Well, that's *their* argument but lots of people think otherwise.  There 
> > appears to be no shortage of people for whom pm-utils does NOT work yet 
> > s2ram does.  https://bugs.launchpad.net/ubuntu/+source/uswsusp/+bug/134238
> 
> The right answer to "This piece of software contains bugs" is not 
> "Provide two pieces of software with the same features but different 
> bugs". The only real functional difference between the two is that

Yes, and that's why pm-utils should die: they have design problems
(depend on hal, can't be pagelocked, unusable for suspend debugging).
									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Justin P. Mattock]

Pavel Machek wrote:
> On Tue 2008-12-16 20:53:35, Matthew Garrett wrote:
>   
>> On Tue, Dec 16, 2008 at 06:07:15PM +0000, Brian J. Murrell wrote:
>>     
>>> On Tue, 16 Dec 2008 16:37:46 +0000, Matthew Garrett wrote:
>>>       
>>>> I seem to be missing the original message, but: no, s2ram doesn't exist
>>>> in Ubuntu because pm-utils handles that role.
>>>>         
>>> Well, that's *their* argument but lots of people think otherwise.  There 
>>> appears to be no shortage of people for whom pm-utils does NOT work yet 
>>> s2ram does.  https://bugs.launchpad.net/ubuntu/+source/uswsusp/+bug/134238
>>>       
>> The right answer to "This piece of software contains bugs" is not 
>> "Provide two pieces of software with the same features but different 
>> bugs". The only real functional difference between the two is that
>>     
>
> Yes, and that's why pm-utils should die: they have design problems
> (depend on hal, can't be pagelocked, unusable for suspend debugging).
> 									Pavel
>   
When using ubuntu intrepid, s2ram
did work, but instead of grabbing
the package from ubuntu, I used Debian
SID instead. (I have a tendency of mixing
packages);

regards;

Justin P. Mattock

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Rafael J. Wysocki]

On Friday, 19 of December 2008, Brian J. Murrell wrote:
> On Thu, 18 Dec 2008 04:24:44 +0000, Brian J. Murrell wrote:
> 
> > 
> > OK.  I've done my s2ram experiments and go no better than with pm-utils/
> > kernel mode suspend (where first suspend/resume works, second suspend
> > works, but second resume hangs the machine).  Here's the full array of
> > suggestions on http://en.opensuse.org/S2ram and their results:
> 
> Just wondering if any of this experimentation is likely to help me or is 
> this just a loser of a laptop that will be more trouble that it's worth 
> trying to get to suspend?  I can take it back to the store still.  Maybe 
> I should and try a different one?
> 
> Or does it seem like there still might be a relatively easy fix for this 
> problem, which would be nice given that it was a pretty good deal (or was 
> it?).  :-)

You're not the only person having a problem like this and unfortunately it's
not easy to debug.  I don't know of any fixes available.

Thanks,
Rafael

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Sergio Monteiro Basto]

[-- Attachment #1: Type: text/plain, Size: 444 bytes --]

On Mon, 2008-12-15 at 17:00 +0000, Brian J. Murrell wrote:
> i915                   64004  2 
> drm                    93736  3 i915

Well , I am thinking if you have a intel-video , and yes ! , I am
getting in a same kind of trouble, I bet that is a problem with i915
drm , mesa 7.2 and xorg-x11-drv-intel (or i810) .
Maybe you can try resume/suspend using vesa video drive, or without
X.org running. 

Regards,
-- 
Sérgio M. B.

[-- Attachment #2: smime.p7s --]
[-- Type: application/x-pkcs7-signature, Size: 2192 bytes --]

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Brian J. Murrell]

[-- Attachment #1: Type: text/plain, Size: 587 bytes --]

On Sun, 2008-12-14 at 07:35 -0800, Justin P. Mattock wrote:
> Is there quite possibly another
> Graphics module you can plug-in,
> Just to see

By module, do you mean software or hardware?  I have to assume you mean
software as this is a laptop.  But I don't know what other module I
could use.  The graphics chip is a:

00:02.1 Display controller: Intel Corporation Mobile GM965/GL960
Integrated Graphics Controller (rev 03)

Maybe I'm just completely mis-understanding you.

BTW, FWIW, this problem happens without an X server even.  Just plain
console driver.

b.


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 197 bytes --]

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Justin P. Mattock]

Is there quite possibly another
Graphics module you can plug-in,
Just to see

justin P. Mattock



On Dec 14, 2008, at 6:08 AM, "Brian J. Murrell"  
<brian@interlinx.bc.ca> wrote:

> On Sun, 2008-12-14 at 03:16 -0600, Robby Workman wrote:
>> In linux.kernel, Brian J. Murrell wrote:
>>> I'm trying to debug a resume problem on a compaq presario C700  
>>> laptop
>>> on the Ubuntu 2.6.24-22-generic kernel.
>>>
>>> SNIPPED
>>>
>>> These symptoms mirror a lot of other reports with this machine in  
>>> that
>>> the first suspend/resume cycle works but a second attempts
>>> (apparently) suspends fine but fails on resume.  The backlight on  
>>> the
>>> video does not come back and neither does the wireless connection,  
>>> so
>>> I'm not positive the problems are purely video related.
>>>
>>> How can I proceed from here to further debug what kind of  
>>> workarounds
>>> or solutions I might be able to use to get reliable suspend/resume
>>> functionality?
>>
>>
>> Assuming Ubuntu uses pm-utils for suspend/hibernate operations,  
>> then I
>> suspect you're needing this patch - the symptoms fit (suspend works
>> the first time, but fails thereafter), especially if a reboot causes
>> suspend to work fine again the first time...
>>
>> http://cgit.freedesktop.org/pm-utils/commit/?id=6c9f2959a36e608e4d1f71230a9eaaa77940c54e
>
> I wish (and was hoping) it was that simple.  The problem is the nature
> of that patch really has no effect on my system because $acpi_flag  
> == 0
> when the sysctl on line 147 is applied.  Presumably this patch has  
> more
> effect if $acpi_flag != 0, by nature of either $QUIRK_S3_BIOS or
> $QUIRK_S3_MODE being == true, neither of which is in my case.  Maybe  
> one
> or both should be.  I dunno.  I'm not sure how these "video quirks"
> work.
>
> b.
>

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Brian J. Murrell]

[-- Attachment #1: Type: text/plain, Size: 1580 bytes --]

On Sun, 2008-12-14 at 03:16 -0600, Robby Workman wrote:
> In linux.kernel, Brian J. Murrell wrote:
> > I'm trying to debug a resume problem on a compaq presario C700 laptop
> > on the Ubuntu 2.6.24-22-generic kernel.
> >
> > SNIPPED
> >
> > These symptoms mirror a lot of other reports with this machine in that
> > the first suspend/resume cycle works but a second attempts
> > (apparently) suspends fine but fails on resume.  The backlight on the
> > video does not come back and neither does the wireless connection, so
> > I'm not positive the problems are purely video related.
> >
> > How can I proceed from here to further debug what kind of workarounds
> > or solutions I might be able to use to get reliable suspend/resume
> > functionality?
> 
> 
> Assuming Ubuntu uses pm-utils for suspend/hibernate operations, then I
> suspect you're needing this patch - the symptoms fit (suspend works
> the first time, but fails thereafter), especially if a reboot causes
> suspend to work fine again the first time...
> 
> http://cgit.freedesktop.org/pm-utils/commit/?id=6c9f2959a36e608e4d1f71230a9eaaa77940c54e

I wish (and was hoping) it was that simple.  The problem is the nature
of that patch really has no effect on my system because $acpi_flag == 0
when the sysctl on line 147 is applied.  Presumably this patch has more
effect if $acpi_flag != 0, by nature of either $QUIRK_S3_BIOS or
$QUIRK_S3_MODE being == true, neither of which is in my case.  Maybe one
or both should be.  I dunno.  I'm not sure how these "video quirks"
work.

b.


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 197 bytes --]

Re: suspend/resume fails on second attempt in LNXVIDEO:00

[Robby Workman]

[-- Attachment #1: Type: text/plain, Size: 1040 bytes --]

In linux.kernel, Brian J. Murrell wrote:
> I'm trying to debug a resume problem on a compaq presario C700 laptop
> on the Ubuntu 2.6.24-22-generic kernel.
>
> SNIPPED
>
> These symptoms mirror a lot of other reports with this machine in that
> the first suspend/resume cycle works but a second attempts
> (apparently) suspends fine but fails on resume.  The backlight on the
> video does not come back and neither does the wireless connection, so
> I'm not positive the problems are purely video related.
>
> How can I proceed from here to further debug what kind of workarounds
> or solutions I might be able to use to get reliable suspend/resume
> functionality?


Assuming Ubuntu uses pm-utils for suspend/hibernate operations, then I
suspect you're needing this patch - the symptoms fit (suspend works
the first time, but fails thereafter), especially if a reboot causes
suspend to work fine again the first time...

http://cgit.freedesktop.org/pm-utils/commit/?id=6c9f2959a36e608e4d1f71230a9eaaa77940c54e

-RW

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 197 bytes --]